1. health and fitness
2. disease

CLINICAL SCIENCES
One-Year, Randomized Study Comparing
Bimatoprost and Timolol in Glaucoma
and Ocular Hypertension
Eve J. Higginbotham, MD; Joel S. Schuman, MD; Ivan Goldberg, MBBS, FRANZCO; Ronald L. Gross, MD;
Amanda M. VanDenburgh, PhD; Kuankuan Chen, MS; Scott M. Whitcup, MD; for the Bimatoprost Study Groups 1 and 2
Objective: To compare bimatoprost with timolol ma-
leate in patients with glaucoma or ocular hypertension.
Methods: In 2 identical, multicenter, randomized,
double-masked, 1-year clinical trials, patients were treated
with 0.03% bimatoprost once daily (QD) (n=474), 0.03%
bimatoprost twice daily (BID) (n = 483), or 0.5% timolol
maleate BID (n = 241).
Main Outcome Measures: Diurnal intraocular pressure (IOP) at 8 AM, 10 AM, and 4 PM and safety variables
(IOP was also measured at 8 PM at selected sites).
BID at most time points, but the efficacy was not as good
as that of the QD regimen. At 10 AM (peak timolol effect)
at month 12, the mean reduction in IOP from baseline
was 7.6 mm Hg (30%) with bimatoprost and 5.3 mm Hg
(21%) with timolol (P⬍.001). A significantly higher
percentage of patients receiving bimatoprost QD (58%)
than timolol (37%) achieved IOPs at or below 17 mm Hg
(10 AM, month 12; P⬍.001). The most common adverse
effect with bimatoprost was hyperemia (significantly
higher with bimatoprost QD than timolol; P⬍.001).
Conclusions: Bimatoprost QD provides sustained IOP
Results: Bimatoprost QD provided significantly lower
lowering superior to timolol or bimatoprost BID and
achieves low target IOPs in significantly more patients.
mean IOP than timolol at every time of the day at each
study visit (P⬍.001). This was also true for bimatoprost
Arch Ophthalmol. 2002;120:1286-1293
From the Department of
Ophthalmology, University of
Maryland at Baltimore
(Dr Higginbotham); the New
England Eye Center, New
England Medical Center, Tufts
University School of Medicine,
Boston, Mass (Dr Schuman);
the Eye Associates and Sydney
Eye Hospital, Sydney, New
South Wales (Dr Goldberg); the
Department of Ophthalmology,
Baylor College of Medicine,
Houston, Tex (Dr Gross); and
Allergan Inc, Irvine, Calif
(Drs VanDenburgh and
Whitcup and Mr Chen). A
complete list of investigators of
the Bimatoprost Study Groups 1
and 2 appears on page 1292.
Drs Higginbotham, Schuman,
Goldberg, Gross and the
members of the study groups
were paid evaluators and do not
have any financial or
proprietary interest in any of
the drugs used in the study
or in the study sponsor.
B
IMATOPROST IS a synthetic
prostamide analogue that
potently lowers intraocular
pressure (IOP).1,2 Bimatoprost has been shown to
help a substantially greater percentage of
patients to achieve low IOPs (ⱕ17 mm Hg)
than does timolol maleate.3,4
Two large-scale, double-masked,
phase 3 clinical (pivotal) trials were
undertaken to evaluate the safety and efficacy of once- (QD) or twice-daily (BID)
regimens of 0.03% bimatoprost ophthalmic solution (Lumigan; Allergan Inc, Irvine, Calif) compared with timolol maleate BID in patients with glaucoma or ocular
hypertension. Interim safety and efficacy
results were evaluated at 3 and 6 months
without breaking study masking for the patients or the study investigators.3,4 These
analyses demonstrated that bimatoprost
QD was more effective in lowering IOP
than was timolol throughout the day at all
study visits through 6 months (Pⱕ.001).
Moreover, a significantly higher percentage of patients treated with bimatoprost
QD achieved low pressures (ⱕ17 mm Hg)
than did patients treated with timolol (the
(REPRINTED) ARCH OPHTHALMOL / VOL 120, OCT 2002
1286
highest P value for the comparison to timolol was .007). Bimatoprost BID was also
more effective than timolol at most time
points, but was slightly less effective than
the QD regimen.
Because glaucoma is a chronic disease, it is important to evaluate the longterm efficacy and safety of new glaucoma
medications. The present report describes the pooled 12-month results from
the 2 pivotal clinical trials comparing bimatoprost with timolol.
METHODS
STUDY DESIGN
Two multicenter, randomized, doublemasked, parallel-group, active-controlled trials were conducted to compare the safety and
efficacy of bimatoprost QD, bimatoprost BID,
and timolol maleate BID. The study protocols
were identical, and the data were pooled for
analysis.
Both studies were conducted in accordance with the Declaration of Helsinki and the
guidelines set forth by the International Council on Harmonization of Technical Requirements for Registration of Pharmaceuticals for
WWW.ARCHOPHTHALMOL.COM
©2002 American Medical Association. All rights reserved.
Downloaded From: http://archfaci.jamanetwork.com/ on 10/15/2014
Human Use and the United States Code of Federal Regulations CFR21. All investigators obtained appropriate institutional review board or ethics committee approval before initiating the study, and all patients provided written informed
consent before any study-related procedures or changes in treatment.
PATIENTS
Key inclusion criteria included a diagnosis of primary openangle glaucoma, chronic angle-closure glaucoma with patent
iridotomy, pseudoexfoliative glaucoma, pigmentary glaucoma, or ocular hypertension. Washout of ocular hypotensive
medications occurred before the baseline visit, and patients were
required to have a postwashout IOP of at least 22 mm Hg and
no greater than 34 mm Hg and a best-corrected visual acuity
of 20/100 or better in each eye. Washout periods ranged from
4 days to 4 weeks, depending on the medication. Parasympathomimetics and carbonic anhydrase inhibitors were washed
out for 4 days, sympathomimetics and topical ␣-agonists were
washed out for 2 weeks, and topical ␤-blockers (alone or in
combination) and topical prostaglandins were washed out for
4 weeks. Key exclusion criteria included any contraindication
to topical ␤-blocker therapy, functionally significant visual field
loss within the past year, filtering surgery within the past 6
months, or other intraocular surgery within the past 3 months.
Women who were pregnant, nursing, planning a pregnancy,
or of childbearing potential and not using a reliable form of contraception were also excluded.
TREATMENT ASSIGNMENT AND DRUG
ADMINISTRATION
At baseline, patients were randomly assigned in a 2:2:1 ratio
to receive 0.03% bimatoprost QD, 0.03% bimatoprost BID, or
0.5% timolol maleate BID. The study sponsor generated the allocation sequence. The bimatoprost QD group received bimatoprost in the evening and a vehicle control solution in the morning to maintain the study mask. Medications were supplied in
identical-appearing bottles that were color coded for use in the
morning or the evening. Patients were instructed to self-instill
their medication into both eyes at approximately 8 AM and 8
PM. On study visits, the morning dose was administered by the
study investigator immediately after the first examination (at
approximately 8 AM). Scheduled visits occurred before the study,
at baseline (day 0), weeks 2 and 6, and months 3, 6, 9, and 12.
EFFICACY AND SAFETY VARIABLES
The primary outcome measure was diurnal IOP. Measurements were performed at 8 AM, immediately preceding the instillation of the morning dose of the study medication, and at
10 AM and 4 PM. Patients at selected sites (approximately 25%
of all enrolled) also had IOP recorded at 8 PM (bimatoprost QD
group, n = 124; bimatoprost BID group, n = 123; and timolol
group, n=65). At month 9, IOP was recorded only at 8 AM and
(at selected sites) 8 PM.
Safety measures included adverse events, visual acuity, visual fields, blood pressure, heart rate, blood chemistry, iris pigmentation, and results of hematology, urinalysis, laser flare photometry, biomicroscopy, and ophthalmoscopy readings. The
severity of adverse events was assessed using the following definitions as guidelines: mild indicates awareness of a sign or a
symptom, but easily tolerated; moderate, enough discomfort
to cause interference with usual activities; and severe, incapacitating, with the inability to work or to perform usual activities. Hyperemia was evaluated on the following scale: none,
0; trace, 0.5; mild, 1.0; moderate, 2.0; and severe, 3.0. To evalu-
(REPRINTED) ARCH OPHTHALMOL / VOL 120, OCT 2002
1287
ate changes in iris pigmentation, investigator(s) at each site compared Polaroid photographs (Polaroid Corporation, Cambridge, Mass) of each patient’s eyes from the baseline and
follow-up visits. A color-calibration strip was photographed beside the eye to verify consistent photographic color processing. A reading center at the sponsor site (consisting of 2 evaluators) also evaluated all photographs in a masked fashion and
obtained slightly lower but generally comparable results. At selected sites, laser-flare photometry readings (using a Kowa FM500 photometer; Kowa Company, Ltd, Chuo-ku, Tokyo, Japan) were recorded before fluorescein instillation or pupil
dilation.
ANALYSES
All randomized patients were included in the efficacy analysis
(intent-to-treat population), and for patients who discontinued before the month 12 visit, the last observed data were carried forward in the analysis for all subsequent time points. An
intent-to-treat analysis with the last observation carried forward is the standard analysis for clinical studies of this type
and is considered the most conservative analysis. It is much
more difficult for a drug to appear to perform well in an intentto-treat analysis with the last observation carried forward than
in a per-protocol analysis because data from patients who did
not receive study medications or who left the study early for
any reason (including lack of efficacy) were kept in the efficacy analysis. Similar results (not shown) were obtained when
the analysis used the per-protocol patient population without
the last observations carried forward. Analysis of IOP used only
data from the eye with the higher IOP at 8 AM at baseline (worse
eye). If both eyes had the same IOP at baseline, the protocol
required that the right eye be used. All patients who received
at least 1 dose of study medication were included in the safety
analyses.
Nominal categorical variables were analyzed using the
Fisher exact test, the Pearson ␹2 test, or Cochran-MantelHaenszel methods.5 Ordinal categorical variables were analyzed by means of the Wilcoxon rank sum test.6 Continuous
variables were analyzed using analysis of variance (ANOVA).
We compared the frequency distributions of patients who had
achieved desirable target IOP levels between groups. We performed the analysis for each target IOP at a given time point
using the Pearson ␹2 test.
Tests of noninferiority and superiority were performed for
the pairwise between-group comparisons of IOP at each time
point. The ␣ level for statistical significance was .05. Noninferiority of bimatoprost was claimed when the upper limit of
the 95% confidence interval of the difference (bimatoprost minus timolol) was no greater than 1 mm Hg. Superiority was
claimed when the upper limit of the confidence interval was
less than 0 mm Hg. The power of each study was at least 0.85
to claim noninferiority of bimatoprost to timolol, based on a
maximum difference of 1.5 mm Hg and using an estimate of
variability (SD=4.052) from a prior study.7 To test for an interaction of drug effect with patient race, the ANOVA was performed with the main effects of treatment group, race (black
vs nonblack), and treatment-by-race interaction. The ␣ level
of significance for the interaction was set at .10 to accommodate a possible low power for this test.
We used the SAS computer program package (version 6.12
and version 7 on Unix; SAS Institute Inc, Cary, NC) for computation and analysis. All the variables were analyzed using SAS
procedures (version 6.12 on Unix; SAS Institute Inc) with the
exception of variables of adverse events, which were analyzed
using version 8.
The sample-size estimate was based on the mean change
in IOP from baseline. With 200 patients in each of the bima-
WWW.ARCHOPHTHALMOL.COM
©2002 American Medical Association. All rights reserved.
Downloaded From: http://archfaci.jamanetwork.com/ on 10/15/2014
Enrolled (N = 1198)
Randomized to
Bimatoprost QD
(n = 474)
Randomized to
Bimatoprost BID
(n = 483)
Randomized to
Timolol Maleate BID
(n = 241)
Completed (n = 415)
Discontinued (n = 59)
Completed (n = 380)
Discontinued (n = 103)
Completed (n = 214)
Discontinued (n = 27)
Figure 1. Trial profile. QD indicates once daily; BID, twice daily.
Table 1. Patient Disposition*
Treatment Groups, No. (%)
Patients
Completed
Discontinued
Lack of efficacy
Adverse events†
Ocular
Nonocular
Protocol violations
Administrative
Other
Bimatoprost
QD
(n = 474)
Bimatoprost
BID
(n = 483)
Timolol
Maleate
(n = 241)
415 (87.6)
59 (12.4)
5 (1.1)
40 (8.4)
25 (5.3)
19 (4.0)
2 (0.4)
11 (2.3)
1 (0.2)
380 (78.7)
103 (21.3)
12 (2.5)
71 (14.7)
60 (12.4)
19 (3.9)
1 (0.2)
18 (3.7)
1 (0.2)
214 (88.8)
27 (11.2)
9 (3.7)
12 (5.0)
4 (1.7)
8 (3.3)
0
5 (2.1)
1 (0.4)
*QD indicates once daily; BID, twice daily.
†Four patients in bimatoprost QD group, 8 in the bimatoprost BID group,
and 0 in the timolol group discontinued owing to both ocular and nonocular
adverse events.
toprost treatment groups and 100 patients in the timolol
group, the power was 0.85 to claim that the mean IOP change
from baseline with bimatoprost QD or BID was more than 1.5
mm Hg greater than that with timolol.
RESULTS
PATIENT POPULATION
A total of 1198 patients were enrolled in the 2 studies.
Of these, 474 patients received bimatoprost QD, 483 received bimatoprost BID, and 241 received timolol at 61
study sites. Fifty sites were located in the United States,
4 in Canada, 5 in Australia, and 2 in New Zealand. The
first patient was enrolled on November 3, 1998, and the
last patient completed the 12 months of treatment on August 4, 2000. One thousand nine patients (84% of all enrolled) completed the study. Of the 189 patients who exited the study early, 123 discontinued owing to (mostly
local) adverse events; 26, owing to lack of efficacy; 34,
for administrative reasons; and 6, for other reasons. Patients were queried at each follow-up visit regarding use
of study medication, and any changes in the study regimen were recorded. Patients or study visits could be excluded from the per-protocol analysis if significant changes
in the study medication regimen were present. However, the present report is based on an analysis of the intent-to-treat population and includes all patients. Full details of patient flow through the study and exit status are
given in Figure 1 and Table 1.
(REPRINTED) ARCH OPHTHALMOL / VOL 120, OCT 2002
1288
We found no statistically significant among-group
differences in demographic characteristics, clinical diagnosis, or baseline IOP (Table 2). We also found no
among-group difference in medical history or the use of
concomitant medication that could affect IOP (eg,
␤-blockers). Medical histories did not designate any patients as low responders to ␤-blockers or prostaglandin
agonists. Most patients were white, with brown or blue
irides and a diagnosis of glaucoma. Approximately 18%
of the study population was black. None of the black patients were from the Australian or New Zealand sites, and
it is consequently unlikely that any Australian aborigines (who have a very low incidence of glaucoma) were
included in the black subpopulation.
IOP-LOWERING EFFICACY
Baseline IOP was similar among the treatment groups
throughout the day (Table 2). Bimatoprost QD provided IOP control that was superior to that of timolol at
all time points and all follow-up visits throughout the 12month study. During all times of the day at which posttreatment measurements were obtained, mean IOP ranged
from 16.2 to 18.0 mm Hg with bimatoprost QD and from
18.2 to 20.0 mm Hg with timolol. As has been reported
previously, bimatoprost BID did not perform as well as
bimatoprost QD.3,4 The presentation of the efficacy results will focus on the data from the bimatoprost QD
group, because this was the study regimen approved by
regulatory authorities.
The peak effect for timolol typically occurs 2 hours
after dosing. Therefore, in this report, the comparative response at each study visit will focus on the 10 AM IOP measurement on each follow-up visit (dosing occurred after the
8 AM examination at each visit). At 10 AM, mean±SEM IOP
ranged from 16.4±0.2 to 17.0±0.2 mm Hg with bimatoprost QD and from 18.2±0.2 to 19.0±0.2 mm Hg with timolol (Figure 2). The difference between the bimatoprost
QD and timolol groups was statistically significant at every follow-up visit (P⬍.001). Mean IOP reductions from
baseline were also significantly greater with bimatoprost
QD than with timolol at this morning peak throughout the
12-month treatment period (P⬍.001). Mean IOP reductions ranged from 7.6 to 8.3 mm Hg (30.2%-32.9%) in the
bimatoprost QD group and from 5.1 to 5.8 mm Hg (20.4%23.3%) in the timolol group. The mean IOP reduction from
baseline consistently ranged from 1.8 to 2.1 mm Hg greater
in the bimatoprost QD group than in the timolol group at
the 10 AM measurement. The decrease from baseline IOP
was up to 8.8 mm Hg in the bimatoprost QD group compared with 6.5 mm Hg in the timolol group (across all times
of day and all study visits).
Significantly higher percentages of patients achieved
low IOP levels with bimatoprost QD than with timolol
(Figure 3). The 10 AM measurement at month 12 revealed that target IOPs at or below 17 mm Hg were
achieved by 57.6% of the bimatoprost QD group compared with 36.5% of the timolol group (P⬍.001). Pressures at or below 15 mm Hg were achieved by 30.6% of
the bimatoprost QD group and 15.8% of the timolol group
(P⬍.001). Patients in the bimatoprost QD group were
up to 21⁄2 times more likely than those in the timolol group
WWW.ARCHOPHTHALMOL.COM
©2002 American Medical Association. All rights reserved.
Downloaded From: http://archfaci.jamanetwork.com/ on 10/15/2014
26
Table 2. Patient Characteristics*
Bimatoprost QD
Timolol Maleate BID
Bimatoprost BID
24
Timolol
Maleate
(n = 241)
P
Value
61.7 ± 12.5 61.6 ± 12.0
61.0 ± 11.4
.74
206 (43.5)
268 (56.5)
234 (48.4)
249 (51.6)
101 (41.9)
140 (58.1)
363 (76.6)
86 (18.1)
6 (1.3)
19 (4.0)
0
371 (76.8)
82 (17.0)
13 (2.7)
15 (3.1)
2 (0.4)
173 (71.8)
46 (19.1)
6 (2.5)
15 (6.2)
1 (0.4)
126 (26.6)
168 (35.4)
14 (3.0)
48 (10.1)
60 (12.7)
5 (1.1)
21 (4.4)
13 (2.7)
13 (2.7)
6 (1.3)
104 (21.5)
173 (35.8)
16 (3.3)
50 (10.4)
62 (12.8)
6 (1.2)
25 (5.2)
16 (3.3)
23 (4.8)
8 (1.7)
49 (20.3)
95 (39.4)
16 (6.6)
26 (10.8)
22 (9.1)
1 (0.4)
10 (4.1)
10 (4.1)
5 (2.1)
7 (2.9)
.48†
270 (57.0)
196 (41.4)
8 (1.7)
270 (55.9)
200 (41.4)
13 (2.7)
133 (55.2)
106 (44.0)
2 (0.8)
.47
300 (63.3)
174 (36.7)
26.0 ± 3.2
24.7 ± 3.7
23.8 ± 3.9
22.1 ± 3.8
323 (66.9)
160 (33.1)
25.9 ± 3.1
24.6 ± 3.7
23.7 ± 4.0
22.2 ± 4.1
147 (61.0)
94 (39.0)
25.8 ± 3.0
24.1 ± 3.5
23.2 ± 3.9
22.4 ± 4.4
.16
22
20
18
†
∗
16
∗
∗
∗
∗
14
12
.77
2
0
4
6
8
10
12
Months of Treatment
Figure 2. Mean ± SEM intraocular pressure (IOP) at 10 AM (peak timolol
effect) at each study visit. Asterisk indicates P⬍.001 vs timolol and
bimatoprost twice daily (BID); dagger, P = .001 vs timolol. QD indicates
once daily.
80
Bimatoprost QD
Timolol Maleate BID
70
.25
.81
.09
.17
.62
Patients Reaching Target IOP, %
Age, mean ± SD, y
Sex
Male
Female
Race
White
Black
Asian
Hispanic
Other
Iris color
Blue
Brown
Green
Dark brown
Yellow-brown
Gray
Blue-gray
Green-brown
Blue/gray-brown
Other
Ophthalmic diagnosis
Glaucoma
OHT
Glaucoma and OHT‡
Washout required
Yes
No
IOP at baseline,
mean ± SD, mm Hg
8 AM
10 AM
4 PM
8 PM§
Bimatoprost Bimatoprost
QD
BID
(n = 474}
(n = 483)
Mean ± SEM IOP, mm Hg
Treatment Groups
∗
69.2
∗
57.6
60
∗
50
47.3
46.6
40
36.5
∗
30.6
30
25.7
∗
20.7
20
10
∗
8.7
6.5
2.1
0
15.8
11.6
∗
≤12
4.6
≤13
≤14
≤15
≤16
≤17
≤18
IOP, mm Hg
*Unless otherwise indicated, data are given as number (percentage).
Percentages have been rounded and may not sum 100. QD indicates once daily;
BID, twice daily; OHT, ocular hypertension; and IOP, intraocular pressure.
†The between-group statistical comparison for iris color is for the
distribution of dark vs light irides.
‡Indicates one eye with glaucoma and the fellow eye with OHT.
§Measurements were obtained at 8 PM at selected sites only (bimatoprost QD
group, n = 124; bimatoprost BID group, n = 123; timolol group, n = 65).
to achieve low pressures that were at or below 12, 13,
14, 15, 16, or 17 mm Hg.
Bimatoprost QD was also superior to timolol in maintaining a low IOP throughout the day. As mentioned in
the two preceding paragraphs, bimatoprost QD produced significantly lower mean IOPs than did timolol at
every time of the day at each follow-up visit throughout
the 12-month treatment period (Pⱕ.001). The overall diurnal mean IOP (average of all measurements at a given
time from all follow-up visits) was significantly lower in
the bimatoprost QD group than the timolol group at each
time of day. Throughout the day, overall mean IOP in
the bimatoprost QD group was approximately 2 mm lower
than the overall mean IOP in the timolol group (P⬍.001;
Figure 4). The IOP-lowering efficacy of the single
evening dose of bimatoprost was still profound 24 hours
after dosing. At the 8 PM measurements, mean±SEM IOP
(REPRINTED) ARCH OPHTHALMOL / VOL 120, OCT 2002
1289
Figure 3. The percentage of patients achieving specific low intraocular
pressures (IOPs) at 10 AM (peak timolol effect) after 12 months of therapy.
Asterisk indicates Pⱕ.001 vs timolol. QD indicates once daily; BID, twice
daily.
ranged from 16.2±0.3 to 17.2±0.3 mm Hg with bimatoprost QD (n=124) compared with 18.2±0.5 to 20.0±0.6
mm Hg with timolol BID (n=65).
A subgroup analysis by race demonstrated that bimatoprost was significantly more effective than timolol
in lowering IOP in both black and nonblack patients. Bimatoprost lowered IOP to the same extent in both black
and nonblack patients, while timolol was notably less effective in blacks than in nonblacks (by up to approximately 2 mm). The statistical validity of this was identified from the significant treatment-by-race interaction
based on the analysis of variance model for repeated measures with the fixed effects of treatment, time, race (black
and nonblack), and treatment-by-race interaction. When
either mean IOP or mean change from baseline IOP was
analyzed for each time of day across all study visits the
difference between blacks and nonblacks was statistically significant at P =.01.
For each analysis of pooled data from both phase 3
trials, similar results were obtained when data from the
WWW.ARCHOPHTHALMOL.COM
©2002 American Medical Association. All rights reserved.
Downloaded From: http://archfaci.jamanetwork.com/ on 10/15/2014
23
Bimatoprost QD
Timolol Maleate BID
Table 3. Treatment-Related Adverse Events*
P
Value,
Bimatoprost Bimatoprost Timolol Bimatoprost
QD
BID
Maleate
QD vs
(n = 474}
(n = 483) (n = 241) Timolol
Treatment Groups, No. (%)
Mean IOP, mm Hg
21
Event
19
∗
17
∗
∗
∗
15
8 AM
4 PM
10 AM
8 PM
Time of Day
Figure 4. Overall diurnal mean intraocular pressure (IOP). The overall mean
diurnal IOP is the average of all measurements at a given time from all
follow-up visits. Asterisk indicates P⬍.001 vs timolol. QD indicates once
daily; BID, twice daily.
25
A
Bimatoprost QD
Timolol Maleate BID
Mean ± SEM IOP, mm Hg
21
19
∗
17
∗
∗
15
B
Bimatoprost QD
Timolol Maleate BID
23
Mean ± SEM IOP, mm Hg
212 (44.7)
271 (56.1)
32 (13.3)
⬍.001
202 (42.6)
69 (14.6)
38 (8.0)
33 (7.0)
259 (53.6)
85 (17.6)
56 (11.6)
32 (6.6)
12 (5.0)
8 (3.3)
5 (2.1)
25 (10.4)
⬍.001
⬍.001
.002
.17
26 (5.5)
26 (5.5)
55 (11.4)
48 (9.9)
1 (0.4)
5 (2.1)
⬍.001
.03
24 (5.1)
24 (5.1)
45 (9.3)
37 (7.7)
8 (3.3)
11 (4.6)
.29
.14
*QD indicates once daily; BID, twice daily.
with bimatoprost QD than with timolol throughout the
day (Figure 5).
SAFETY AND TOLERABILITY
23
25
Conjunctival
hyperemia
Eyelash growth
Eye pruritus
Eye dryness
Burning sensation
in the eye
Eyelid pigmentation
Foreign body
sensation
Eye pain
Visual disturbance
21
19
∗
17
∗
∗
15
8 AM
10 AM
4 PM
8 PM
Time of Day at Month 12
Figure 5. Diurnal mean intraocular pressure (IOP) at month 12 for trials 1 (A)
and 2 (B) separately. There was inadequate power for pairwise comparisons at
8 PM because the sample sizes were too small at this time. Asterisk indicates
Pⱕ.001 vs timolol. BID indicates twice daily; QD, once daily.
individual trials were evaluated separately. For example, in each trial and in the pooled data, diurnal mean
IOP at month 12 was consistently 2 to 3 mm Hg lower
(REPRINTED) ARCH OPHTHALMOL / VOL 120, OCT 2002
1290
We found a low rate of discontinuations due to adverse
events. All 3 treatment regimens were safe and well tolerated. The most common adverse effects associated with
bimatoprost were conjunctival hyperemia and eyelash
growth.
Other potentially treatment-related adverse events
that occurred in more than 5% of the bimatoprost QD
group are listed in Table 3 and included eye pruritus,
eye dryness, eye burning, eyelid pigmentation, foreignbody sensation, eye pain, and visual disturbance. For some
of these adverse events, we found a significantly higher
incidence in the bimatoprost QD group than in the timolol group. However, no statistically significant difference was found between the bimatoprost QD and timolol groups in the number of reports of eye pain, visual
disturbance, and burning sensation in the eye. Patients
in the bimatoprost BID group had a higher incidence of
hyperemia, eyelash growth, eyelid pigmentation, and eye
pain than did patients in the bimatoprost QD group
(P⬍.01). Most treatment-related adverse events were mild
in severity.
Assessment of the hyperemia findings were confounded by the fact that trace or greater conjunctival hyperemia was present at baseline (before the initial administration of the study medication) in 25.1% of the
patients in the bimatoprost QD group and in 17.8% of
the patients in the timolol group. Mean scores of conjunctival hyperemia (worse severity of the eyes) were low
(in the trace range) in the bimatoprost QD treatment
group and remained low throughout the study
(Figure 6). Only 5.3% of patients experienced greater
than a 1-U (mild) increase in hyperemia during this 12month study. Only 3.4% of the patients in the bimatoprost QD group, 5.6% of those in the bimatoprost BID
group, and 0.4% of those in the timolol group discontinued from the study owing to hyperemia (P =.01; biWWW.ARCHOPHTHALMOL.COM
©2002 American Medical Association. All rights reserved.
Downloaded From: http://archfaci.jamanetwork.com/ on 10/15/2014
Severe
3.0
16
Bimatoprost QD
Timolol Maleate BID
Mild or Greater Increase
in Hyperemia
Mean Laser-Flare Photometer Reading
2.5
Mean Scores
Moderate
2.0
1.5
Mild
1.0
Trace
0.5
None
0.0
No or Trace Change
in Hyperemia
14
12
10
8
6
4
0
2
4
6
8
10
12
0
3
Months of Treatment
6
9
12
Time, mo
Figure 6. Mean conjunctival hyperemia scores (obtained in the worse eyes).
The scale for evaluating hyperemia is explained in the “Efficacy and Safety
Variables” subsection of the “Methods” section. QD indicates once daily;
BID, twice daily.
Figure 7. Mean laser-flare photometer readings in patients with greater or
lesser changes in hyperemia. Readings are given in photon counts per
millisecond.
matoprost QD vs timolol). The median time to onset of
hyperemia was 14 days after the initiation of therapy, and
in most cases it began within the first 6 weeks. No patients receiving bimatoprost QD discontinued the study
because of conjunctival hyperemia after month 6.
Laser-flare photometry measurements (a common
measure of ocular inflammation) of a subset of 310 patients (124 in the bimatoprost QD group, 123 in the bimatoprost BID group, and 63 in the timolol group) demonstrated that there were no significant differences among
the treatment groups in mean laser-flare photometry readings or in mean changes from baseline. In addition, we
found no increase in flare readings in those patients in
whom conjunctival hyperemia developed. In fact, after
12 months of bimatoprost QD treatment, mean laserflare photometry readings were lower in patients who had
a mild or greater increase in conjunctival hyperemia (7.67
photon counts/ms) compared with those in patients who
had no change, a trace increase, or a decrease in hyperemia (9.26 photon counts/ms) (Figure 7); however, this
difference was not statistically significant (P = .37).
Changes in iris pigmentation, determined by the results of the masked investigators’ evaluation of the photographs, were reported in 7 (1.5%) of the 474 patients
in the bimatoprost QD group, 9 (1.9%) of the 483 patients in the bimatoprost BID group, and no patients in
the timolol group. Generally comparable findings were
obtained by the sponsor’s reading center.
The incidence of adverse events leading to discontinuation was comparably low in the bimatoprost QD and
timolol groups. In the bimatoprost QD treatment group,
8.4% of patients discontinued owing to adverse events,
compared with 5.0% in the timolol group (P = .09). The
discontinuation rate due to ocular adverse events was 5.3%
in the bimatoprost QD group and 1.7% in the timolol
group. This difference was statistically significant (P=.02),
exclusively because of the statistically significant difference in the number of patients who discontinued owing
to hyperemia (3.4% in the bimatoprost QD group and
0.4% in the timolol group; P=.01). We found no betweengroup difference in the number of patients who dis-
continued because of any other adverse event. In the
bimatoprost BID treatment group, 14.7% of patients
discontinued because of adverse events (P = .003 compared with bimatoprost QD).
No clinically significant changes in results of ophthalmoscopy, laser-flare photometry readings, visual acuity, visual fields, or systemic safety variables were found
with any treatment regimen.
(REPRINTED) ARCH OPHTHALMOL / VOL 120, OCT 2002
1291
COMMENT
The pooled 1-year results of the 2 pivotal phase 3 trials
clearly demonstrated that bimatoprost QD was superior
to timolol in long-term IOP reduction. Significantly higher
percentages of patients achieved low pressures with bimatoprost QD, and IOP was controlled throughout the
day. The efficacy of bimatoprost QD was sustained
throughout the 1-year study. The mean reduction from
baseline IOP in the bimatoprost QD group was very consistent (range, 30%-33%) at follow-up visits through 1
year of treatment.
Controlling IOP is critical to slowing the progression of glaucomatous damage in patients with glaucoma
or ocular hypertension. Elevated IOP is the single most
important risk factor for open-angle glaucoma, and numerous studies8-12 have found that reducing IOP decreases the risk for visual field loss in patients with glaucoma. A recent analysis by the Advanced Glaucoma
Intervention Study group demonstrated that after surgery to reduce IOP, patients who consistently had IOPs
less than 18 mm Hg at every clinic visit during the 6-year
follow-up showed mean changes in visual fields close to
0.12 In these patients, the mean IOP was very low (12.3
mm Hg), suggesting that lower IOPs result in greater patient benefit. Reductions in IOP of 30% or greater have
also been shown to significantly reduce the rate of glaucomatous progression in patients with normal-tension
glaucoma.8 Even in patients without evidence of optic disc
cupping, reduction of an elevated IOP can be beneficial
in delaying the onset of early glaucomatous damage.9,10
Therefore, a goal of glaucoma therapy should be the
WWW.ARCHOPHTHALMOL.COM
©2002 American Medical Association. All rights reserved.
Downloaded From: http://archfaci.jamanetwork.com/ on 10/15/2014
achievement and maintenance of target IOPs that are low
enough to minimize glaucomatous progression.13 On the
basis of the superior IOP-lowering efficacy of bimatoprost QD demonstrated in the present study, long-term
bimatoprost QD treatment may reasonably be assumed
to provide greater protection for the visual field than does
timolol in patients with glaucoma or ocular hypertension. The greater ability of bimatoprost to help patients
reach IOP levels more consistently below 18 mm Hg
should prove particularly beneficial.8,12
A recent study by Asrani and colleagues14 reported
that large fluctuations in diurnal IOP are an independent risk factor for the progression of glaucomatous damage. The results of the present study show that bimatoprost QD effectively controls IOP fluctuations throughout
the day that could otherwise increase the risk for further optic nerve damage.15,16
Numerous studies have reported significant differences between black and nonblack populations in the
prevalence of glaucoma17,18 and the response to treatment.18 Population-based studies have reported that glaucoma is 4 to 6 times more prevalent in black than in nonblack populations and that glaucoma is 6 to 8 times more
likely to lead to blindness in black than in nonblack populations.17 The reasons for the greater prevalence and visual field loss in black populations are not yet known.
The results of the present study demonstrated that bimatoprost QD reduces IOP as effectively in black as in
nonblack patients, and that bimatoprost QD provides IOP
lowering superior to that of timolol across both of these
patient populations.
Throughout 1 year of treatment, bimatoprost QD was
safe and well tolerated. The most common adverse effect,
conjunctival hyperemia, was generally well tolerated and
mild in severity. In our experience, hyperemia can resolve after 2 to 4 weeks of treatment. The lack of any correlation between hyperemia and laser-flare photometry
readings indicates that there was no association with intraocular inflammation and confirms the general finding that this adverse effect does not represent a significant clinical safety concern. The investigating physicians
reported an increase in iris pigmentation in only 1.5%
of the patients in the bimatoprost group treated for 12
months. By comparison, in the phase 3 clinical trials of
latanoprost in which iris pigmentation changes were
evaluated at a reading center, the incidence of increased
iris pigmentation ranged from 10.9% to 22.9% of patients treated with latanoprost QD for 12 months.19 These
results suggest that iris pigmentation changes may occur less frequently with bimatoprost than with latanoprost. However, given the different techniques used in
the studies of these 2 drugs and the variation in the ability of investigators to detect iris pigment changes, a longterm comparison of bimatoprost with latanoprost using
a single method will be needed to evaluate this issue more
thoroughly.
Bimatoprost BID was associated with a significantly
higher incidence of certain adverse events than was
bimatoprost QD. This, along with the greater efficacy of
the QD regimen, supports the once-daily use of this
drug.
(REPRINTED) ARCH OPHTHALMOL / VOL 120, OCT 2002
1292
Study Group Investigators
Bimatoprost Study Group 1
Mark B. Abelson, MD (North Andover, Mass); George
Baerveldt, MD (Cleveland, Ohio); Stephen Best, MD (Auckland, New Zealand); James D. Branch, MD (WinstonSalem, NC); James D. Brandt, MD (Sacramento, Calif);
John Brennan, MD (Sherman, Tex); Anne M. V. Brooks,
MD (East Melbourne, Victoria); Salim Butrus, MD (Washington, DC); Leonard R. Cacioppo, MD (Brooksville, Fla);
Guy D’Mellow, MD (Brisbane, Queensland); Harvey B. DuBiner, MD (Morrow, Ga); Efraim Duzman, MD (Irvine,
Calif); Robert J. Foerster, MD (Colorado Springs, Colo);
Jonathan Frantz, MD (Fort Myers, Fla); Walter I. Fried,
MD, PhD (Gurnee, Ill); David K. Gieser, MD (Wheaton,
Ill); Richard A. Lewis, MD (Sacramento, Calif); Andrew
Logan, MD (Wellington, New Zealand); Richard McGovern, MD (Adelaide, South Australia); Thomas Mundorf,
MD (Charlotte, NC); George F. Nardin, MD (Kailua, Hawaii); Jonathan Nussdorf, MD (Louisville, Ky); Julian Rait,
MD (East Melbourne); Robert L. Shields, MD (Denver,
Colo); Thomas R. Walters, MD (Austin, Tex); Jeffrey C.
Whitsett, MD (Houston, Tex); Jacob T. Wilensky, MD
(Chicago, Ill); Robert Williams, MD (Louisville).
Bimatoprost Study Group 2
Allen Beck, MD (Atlanta, Ga); Louis Cantor, MD (Indianapolis, Ind); George Cioffi, MD (Portland, Ore); John S.
Cohen, MD (Cincinnati, Ohio); David Cooke, MD (St Joseph, Mich); Andrew Crichton, MD (Calgary, Alberta);
Denise F. Dudley, MD (Bellingham, Wash); Richard Evans,
MD (San Antonio, Tex); Stephen Greenberg, MD (Holbrook, NY); Neeru Gupta, MD, PhD (Toronto, Ontario);
Leonard Gurevich, MD (West Seneca, NY); Oscar Kasner, MD (Montreal, Quebec); Donald Kellum, MD (Boulder, Colo); Melvyn Koby, MD (Louisville); John Kwedar, MD (Springfield, Ill); David McGarey, MD (Flagstaff,
Ariz); Frederick Mikelberg, MD (Vancouver, British Columbia); Robert Noecker, MD (Tucson, Ariz); Leon L. Remis, MD (Marblehead, Mass); Robert Ritch, MD (New
York, NY); Michael Rotberg, MD (Charlotte); Howard I.
Schenker, MD (Rochester, NY); Elizabeth Sharpe, MD (Mt
Pleasant, SC); Mark B. Sherwood, MD (Gainesville, Fla);
Joseph Sokol, MD (Waterbury, Conn); Alfred Solish, MD
(Pasadena, Calif); Julia Whiteside-Michel, MD (Little Rock,
Ark); and Barbara Wirostko, MD (Huntington Station, NY).
CONCLUSIONS
We found 0.03% bimatoprost QD to be clinically and statistically superior to 0.5% timolol maleate BID in lowering
IOPs in patients with glaucoma or ocular hypertension. Significantly more patients achieve very low target pressures
with bimatoprost QD than with timolol maleate BID. Bimatoprost QD provides potent IOP control throughout the
day and is safe and well tolerated. The efficacy of bimatoprost QD is sustained through at least 1 year of treatment.
Submitted for publication January 31, 2002; final revision
received May 15, 2002; accepted July 5, 2002.
This study was sponsored by Allergan Inc, Irvine, Calif.
This study was presented at the Annual Meeting of the
American Academy of Ophthalmology, New Orleans, La,
November 13, 2001.
WWW.ARCHOPHTHALMOL.COM
©2002 American Medical Association. All rights reserved.
Downloaded From: http://archfaci.jamanetwork.com/ on 10/15/2014
Corresponding author and reprints: Eve J. Higginbotham, MD, Department of Ophthalmology, University of
Maryland at Baltimore, 419 W Redwood St, Suite 580, Baltimore, MD 21201-1595 (e-mail: fcwejh6786@aol.com).
9.
10.
REFERENCES
1. Krauss AH-P, Chen J, Andrews SW, et al. The ocular pharmacology and bioavailability of AGN 192024 (Lumigan™), a novel ocular hypotensive agent [abstract]. Invest Ophthalmol Vis Sci. 2001;42(4, suppl):S832.
2. Woodward DF, Krauss AH, Chen J, et al. The pharmacology of bimatoprost
(Lumigan). Surv Ophthalmol. 2001;45(4, suppl):S337-S345.
3. Brandt JD, VanDenburgh AM, Chen K, Whitcup SM, for the Bimatoprost Study
Group. Comparison of once- or twice-daily bimatoprost with twice-daily timolol
in patients with elevated IOP: a 3-month clinical trial. Ophthalmology. 2001;108:
1023-1031.
4. Sherwood M, Brandt J, for the Bimatoprost Study Groups 1 and 2. Six month
comparison of bimatoprost once-daily and twice-daily with timolol twice-daily
in patients with elevated intraocular pressure. Surv Ophthalmol. 2001;45(suppl
4):S361-S368.
5. Fleiss JL. Statistical Methods for Rates and Proportions. 2nd ed. New York, NY:
John Wiley & Sons Inc; 1981.
6. Lehmann EL. Nonparametrics: Statistical Methods Based on Ranks. San Francisco, Calif: Holden-Day; 1975.
7. Laibovitz RA, VanDenburgh AM, Felix C, et al. Comparison of the ocular hypotensive lipid AGN 192024 with timolol: dosing, efficacy, and safety evaluation of
a novel compound for glaucoma management. Arch Ophthalmol. 2001;119:9941000.
8. Collaborative Normal-Tension Glaucoma Study Group. Comparison of glaucomatous progression between untreated patients with normal-tension glaucoma
and patients with therapeutically reduced intraocular pressures [published cor-
11.
12.
13.
14.
15.
16.
17.
18.
19.
rection appears in Am J Ophthalmol. 1999;127:120]. Am J Ophthalmol. 1998;
126:487-197.
Epstein DL, Krug JH Jr, Hertzmark E, Remis LL, Edelstein DJ. A long-term clinical trial of timolol therapy versus no treatment in the management of glaucoma
suspects. Ophthalmology. 1989;96:1460-1467.
Kass MA, Gordon MO, Hoff MR, et al. Topical timolol administration reduces the
incidence of glaucomatous damage in ocular hypertensive individuals: a randomized, double-masked, long-term clinical trial. Arch Ophthalmol. 1989;107:
1590-1598.
Mao LK, Stewart WC, Shields MB. Correlation between intraocular pressure control and progressive glaucomatous damage in primary open-angle glaucoma. Am
J Ophthalmol. 1991;111:51-55.
The AGIS Investigators. The Advanced Glaucoma Intervention Study (AGIS), VII:
the relationship between control of intraocular pressure and visual field deterioration. Am J Ophthalmol. 2000;130:429-440.
Singh K, Spaeth G, Zimmerman T, Minckler D. Target pressure: glaucomatologists’ holey grail. Ophthalmology. 2000;107:629-630.
Asrani S, Zeimer R, Wilensky J, Gieser D, Vitale S, Lindenmuth K. Large diurnal
fluctuations in intraocular pressure are an independent risk factor in patients with
glaucoma. J Glaucoma. 2000;9:134-142.
Sacca SC, Rolando M, Marletta A, Macri A, Cerqueti P, Ciurlo G. Fluctuations of
intraocular pressure during the day in open-angle glaucoma, normal-tension glaucoma and normal subjects. Ophthalmologica. 1998;212:115-119.
Zeimer RC, Wilensky JT, Gieser DK. Presence and rapid decline of early morning intraocular pressure peaks in glaucoma patients. Ophthalmology. 1990;97:547-550.
Sommer A, Tielsch JM, Katz J, et al. Racial differences in the cause-specific prevalence of blindness in east Baltimore. N Engl J Med. 1991;325:1412-1417.
Sommer A, Tielsch JM, Katz J, et al, and the Baltimore Eye Survey. Relationship
between intraocular pressure and primary open angle glaucoma among white
and black Americans. Arch Ophthalmol. 1991;109:1090-1095.
Wistrand PJ, Stjernschantz J, Olsson K. The incidence and time-course of latanoprost-induced iridial pigmentation as a function of eye color. Surv Ophthalmol. 1997;41(suppl 2):S129-S138.
CME Announcement
CME Hiatus: July Through December 2002
CME from JAMA/Archives will be suspended between July and December
2002. Beginning in early 2003, we will offer a new online CME program that
will provide many enhancements:
•
•
•
•
Article-specific questions
Hypertext links from questions to the relevant content
Online CME questionnaire
Printable CME certificates and ability to access total CME credits
We apologize for the interruption in CME and hope that you will enjoy
the improved online features that will be available in early 2003.
(REPRINTED) ARCH OPHTHALMOL / VOL 120, OCT 2002
1293
WWW.ARCHOPHTHALMOL.COM
©2002 American Medical Association. All rights reserved.
Downloaded From: http://archfaci.jamanetwork.com/ on 10/15/2014